Is a second home battery worth it? A simulation with eight months of minute data
My home battery – a Marstek Venus E with 5.12 kWh – is full by midday on sunny days. After that, the surplus goes to the grid for a few cents. The obvious idea: a second battery of the same kind. Cost: around €850, including the additional smart meter it needs.
Before spending the money, I wanted to know for sure. My energy management system JG Energy stores a daily log with one value per minute: grid import or export, PV power, battery and wallbox. That’s enough to simulate a second battery quite realistically.
The data
- Period: 8 March to 4 November 2026, 242 days
- Resolution: 1 minute, so roughly 350,000 data points
- System: PV with a good 17 kW of inverter power plus two balcony solar systems, one 5.12 kWh battery, a wallbox for two EVs
The crucial point: the first battery is already in the data. The grid meter only shows what’s left after the existing battery has charged and discharged. That remainder is exactly what a second battery would have to work with.
The model
The simulated second battery behaves like the first:
- usable capacity: 88 % of 5.12 kWh (the Marstek only discharges down to about 12 %)
- power: up to 2.5 kW charging and discharging
- efficiency: 92 % in each direction, so about 85 % round trip
- charges only from surplus that would otherwise be exported
- discharges only when importing from the grid – in two variants, matching the “home battery while the car charges” setting in JG Energy: “stays off” (no discharging while the wallbox is charging) and “helps” (the battery also discharges into the car)
- prices: grid electricity around 27 ct/kWh, feed-in tariff just under 8 ct/kWh
Every kWh the second battery delivers to the house saves grid electricity. Every kWh it charges is missing from the feed-in revenue. The core of the simulation is correspondingly short:
for minute in day:
grid = minute.grid_w # + import, - export
if grid < 0: # surplus → charge
e = min(-grid, 2500) / 60 / 1000
e = min(e, (cap - soc) / 0.92)
soc += e * 0.92; charged += e
elif grid > 0 and (helps or minute.wallbox_w == 0): # import → discharge
e = min(grid, 2500) / 60 / 1000
e = min(e, soc * 0.92)
soc -= e / 0.92; discharged += e
savings = discharged * price_grid - charged * price_feed_in
The result
With the “stays off” setting, the second battery would have charged about 400 kWh over the 242 days and delivered 340 kWh. That’s a little over €60.
With “helps” a bit more comes in: the battery can then also discharge into the car and is ready to absorb surplus again the next day. That gives about 490 kWh charged, 410 kWh delivered and roughly €75.
The table shows the “stays off” variant. Grid import and export are the measured values – i.e. what was left after the existing battery.
| Month | Grid import | Export | Savings |
|---|---|---|---|
| March (from 8th) | ~100 kWh | ~540 kWh | ~€9 |
| April | ~50 kWh | ~1,930 kWh | ~€5 |
| May | ~25 kWh | ~2,010 kWh | ~€2.50 |
| June | ~95 kWh | ~1,940 kWh | ~€5 |
| July | ~60 kWh | ~1,900 kWh | ~€5.50 |
| August | ~35 kWh | ~1,810 kWh | ~€3 |
| September | ~145 kWh | ~740 kWh | ~€15 |
| October | ~625 kWh | ~365 kWh | ~€16 |
November to February bring hardly any surplus, so hardly anything for a second battery either. Extrapolated to a full year, I get around €70 (“stays off”) or around €85 (“helps”).
At around €850 purchase price, that’s 10 to 12 years until the battery has paid for itself – without degradation, without the device’s standby consumption, without the cost of capital. That puts the payback period close to the battery’s lifetime.
Why does the second battery deliver so little?
The table makes it clear: in summer there’s surplus galore – almost 2,000 kWh of export per month. The second battery was full at some point on 225 of 242 days. Charging isn’t the problem.
Discharging is. From May to August, only 25 to 100 kWh of grid import per month were left after the first battery. In summer, the first battery already covers the nights almost completely. The second one sits there full and isn’t needed.
In autumn it’s the other way round: in October there was plenty of grid import, but too little surplus to fill two batteries.
So a second battery only pays off on days when both come together: more surplus than the first battery can absorb and more consumption than it can cover. For me, that’s mainly March and September.
That’s the real lesson: the first battery harvests the easy cycles. Every additional kWh of storage brings considerably less than the one before.
What about a bigger second battery?
I ran the same simulation with other sizes (same period, 8 months, “stays off” setting):
| Additional capacity | Savings |
|---|---|
| 2.56 kWh | ~€45 |
| 5.12 kWh | ~€60 |
| 10.24 kWh | ~€70 |
| 15.36 kWh | ~€72 |
Beyond about 10 kWh of additional capacity, practically nothing changes. More storage simply finds no more consumption to discharge into.
At what price would it pay off?
My target would have been a payback period of 3 to 4 years. At around €70 to €85 a year, the complete set – battery plus smart meter – would have to cost only about €200 to €350. Current prices are far from that.
It could become more interesting if consumption changes. A heat pump or an air conditioner running in the evening creates exactly the consumption the second battery is missing. So does a dynamic electricity tariff that lets the battery charge from cheap grid power as well. I deliberately didn’t simulate either of these here.
Conclusion
Depending on the setting, a second battery of the same size would have saved me around €70 to €85 a year and only paid for itself after 10 to 12 years. So I’m not buying it.
If you’re thinking about adding storage yourself: before calculating how much surplus you have, look at how much grid import is left after your existing battery. If that’s small, even the biggest battery can’t save anything. And if you have minute data from your grid meter, you can run the numbers yourself with a few lines of code.